frame.io

[Christoffer Glans]

I'm getting very poor performance playing videos even on a high speed broadband. This is seriously hurting customer relations when sending them offline cuts to approve.

 

[jacob.schwarz]

Same here. I gave up on frame.io hopefully they'll fix it.

 

[Christoffer Glans]

So where does the money go to? They charge pretty high for their service so I would have thought that most of it went to high speed servers?

 

[Terry VerHaar]

I have had minimal use of frame.io so far but have to say it worked fine. I loaded a very "light" file, though. I'd be interested in what you guys uploaded and what kind of performance glitches you are getting. I really had high hopes for it as a possible dailies mechanism. This is disappointing.

 

[Christoffer Glans]

One feature film (1h 40min) at 1,5 gb... didn't seem to work at all. Few 30 second clips, lags in the beginning, impossible to view without the beginning lagging. Music videos around 2min30sec - 4min, plays fine then stops and lags, client is pissed off at this type of client viewing.

The thing is that frame.io is supposed to be aimed at professionals and playback issues is in no way tolerated for such a service. I was hyped in the beginning, but I've had too many clients getting angry so if they don't fix this priority I will never use it again.

 

[Chris McKechnie]

I've had the same issues, with multiple browsers. Frame IO plays back at an incredibly high bit-rate - blowing the doors off of Vimeo in terms of quality. They are working on a playback resolution selector thing which basically would allow people with slower internet connections to adjust the playback resolution and drop it down to a lower bandwith. Their customer service has been unsurpassed though...instantly responding to questions and feedback comments. Really impressed! Remember, Frame IO is in its infancy and it already has impressed me.

 

[Terry VerHaar]

Chris - I agree with you that it has great potential. And I, too, have had near instant response from Emery about any issues. Like Christoffer, I really want to get hooked on it but I am hoping they are on a rapid improvement road. Personally, I am more interested in a dailies solution than a client approval platform so I would love to see extensive metadata capabilities vis a vis Lightiron's LivePlay3.

 

[Björn Benckert]

Not to derail the thread about the playback problems. I had them some times but then log out and back in again solved it. I upload proress422 for shorter things and h.264 with ipad settings for longer things and it all worked quite well.

But the one thing I dont like with frame IO is that when I add collaborators they need to sign in with google account or such. Why is it not possible to just set a general password for the project and let people sign in with that password and their email.

Clients tend to think it´s to much of a hassle to get in as it is now so they ask for a wetransfer or something similar instead...

Or have I missed something is it possible to give them a simple login with common password?

As a lot of times I send a login to an agency producer, then I get a mail back with 15 email adresses that she wants me to add to the list. At those occations it would be better to just send her a user: client pass: client login or such instead of invites to each and every one.

 

[Christoffer Glans]

Björn, problem with the playback is that even if I can play it on my connection it needs to be at youtube/vimeo playback-capabilites for clients. If I send something to a director for approval, like I am at the moment, and the director is in another country on a hotel connection, they still need to be able to play the edit. It's the whole point of an online collaboration site to be able to give that ability to collaborators, otherwise I could just use my FTP and Email or phone and I would still get all the info I need in order to do the edit, while the client won't have any problems with the playback.

And I'm with you on the login, it's a hassle. Frame.io should be a site like Vimeo, easy to register or login, but even if you aren't on Vimeo, a private password for total accessibility to the video and comment section should be a no-brainer. It's easy to spot what software and web applications which survives and thrives today. It all spells "easy to use". Why did a little company like Spotify get almost as much power as Apple in the music industry distribution? Because of the ease of use, just login and play your music, no bullshit.

People don't have time to manage accounts and such, it's not what they do or should do. If you want a tool to survive the long run and be the industry leader in client and collaborator-communications for video and audio reviews it should be the most simple tool available. I often find myself using the tool with the least features because it is most of the time the smoothest and easiest tool available. Every time I need to go onto a forum to fix some login or rather unnecessary error that occurs I say to myself... this won't survive long.

 

[Björn Benckert]

Björn, problem with the playback is that even if I can play it on my connection it needs to be at youtube/vimeo playback-capabilites for clients. If I send something to a director for approval, like I am at the moment, and the director is in another country on a hotel connection, they still need to be able to play the edit. It's the whole point of an online collaboration site to be able to give that ability to collaborators, otherwise I could just use my FTP and Email or phone and I would still get all the info I need in order to do the edit, while the client won't have any problems with the playback.

And I'm with you on the login, it's a hassle. Frame.io should be a site like Vimeo, easy to register or login, but even if you aren't on Vimeo, a private password for total accessibility to the video and comment section should be a no-brainer. It's easy to spot what software and web applications which survives and thrives today. It all spells "easy to use". Why did a little company like Spotify get almost as much power as Apple in the music industry distribution? Because of the ease of use, just login and play your music, no bullshit.

People don't have time to manage accounts and such, it's not what they do or should do. If you want a tool to survive the long run and be the industry leader in client and collaborator-communications for video and audio reviews it should be the most simple tool available. I often find myself using the tool with the least features because it is most of the time the smoothest and easiest tool available. Every time I need to go onto a forum to fix some login or rather unnecessary error that occurs I say to myself... this won't survive long.

Agree,
But I have not had the playback issues you refer to. But I can only assume that frame.io does not have the same backbone as for example vimeo or youtube. To get a global video server to work properly they need to mirror their content to all sub nets, if not they will hit lagg. For example we got a super fast fiber connection at our office. 1000Mb/s both directions but if somebody in the states try to fetch something from our ftp it can be as bad as less than one procent of that speed even if they got a fast connection at their end. Thats due to the fact that all internet providers can only guarantee speed within their sub net. In Sweden that is normally the KTH net that is in use, which normally gives great preformance within sweden or even europe but not transatlantic or down to Australia or such. And I can only think that Frame.io´s problem that you describe is due to the above, simply they need to host their content on all continents / sub nets to be able to provide the speed needed.

But thats just my not so qualified guess

 

[Emery Wells]

Hey everyone, Frame.io co-founder here. First off, sucks to hear many of you are having playback trouble but we are aware of the issue and we're actively working on solutions that will be available in a matter of weeks. As Chris McKechnie pointed out, we're still a brand new product/company and we've seen tremendous growth over the past few months. Building an infrastructure that supports users in over 160 countries has it challenges. Our primary data center is located on the east coast of the United States. The further you are from that data center, the greater your latency and poorer your performance will be. For many international members, this means a pretty crappy playback experience. We do use what's called a content distribution network which means when you request to play a video from a geographically distant location, we instantly move the asset from our united states server to an edge location nearest you. This type of architecture works great for a company like YouTube and a million people in Sweden all the sudden want to watch the latest viral video. It can be cached at an edge location and all future requests will be accessed from that cache. But for a service like Frame.io where videos are being accessed private and infrequently, this solution doesn't work well because it's still slow to move the file from our origin to the cahched location. So what are we going to do:

1. First, we'll offer a lower bit rate resolution (like YouTube). Right now our playback is a 5,000kbps 1080p stream. This is just too high for many connections. Offering lower bitrates will solve the problem for most of you.
2. We're designing a new architecture that can better handle the geographically dispersed users. This means if you live in Sweden you'll likely be uploading to a data center near you where that asset will permanently live. We're also testing other ideas.
3. Upload performance from regions outside the U.S. will get dramatically better in the coming weeks as well.

It's important to note that everything you upload to Frame.io get's converted server side (like YouTube and Vimeo) so it doesn't matter what you upload. That being said, resolution does matter. If you upload 1280 x 720 or lower, we will not up convert is to 1080 so the stream winds up around 2500 kbps. So that's a temporary solution to improving playback performance.

The other issue that was mentioned about it taking a while to start when you press play is a Safari bug we're aware of that we're fixing very soon. These are all just growing pains for a company that's only 4 months old. There are tons of areas we're working in improving but being able to upload and playback are obviously the most central so we're focusing on it now and this will all be a distant memory soon.

Thanks for the continued support guys!

 

[Emery Wells]

Björn, problem with the playback is that even if I can play it on my connection it needs to be at youtube/vimeo playback-capabilites for clients. If I send something to a director for approval, like I am at the moment, and the director is in another country on a hotel connection, they still need to be able to play the edit. It's the whole point of an online collaboration site to be able to give that ability to collaborators, otherwise I could just use my FTP and Email or phone and I would still get all the info I need in order to do the edit, while the client won't have any problems with the playback.

And I'm with you on the login, it's a hassle. Frame.io should be a site like Vimeo, easy to register or login, but even if you aren't on Vimeo, a private password for total accessibility to the video and comment section should be a no-brainer. It's easy to spot what software and web applications which survives and thrives today. It all spells "easy to use". Why did a little company like Spotify get almost as much power as Apple in the music industry distribution? Because of the ease of use, just login and play your music, no bullshit.

People don't have time to manage accounts and such, it's not what they do or should do. If you want a tool to survive the long run and be the industry leader in client and collaborator-communications for video and audio reviews it should be the most simple tool available. I often find myself using the tool with the least features because it is most of the time the smoothest and easiest tool available. Every time I need to go onto a forum to fix some login or rather unnecessary error that occurs I say to myself... this won't survive long.

If it isn't clear from our UX and UI, we obsessively focus on simplicity. Requiring a login was a philosophical choice. Frame.io is trying to fundamentally change the workflow of emailing links. Emailing links is what makes everything suck. It's unorganized and totally impossible to keep track of. Every time you want to access something you have to find the link (and perhaps the associated password) in your email. If this were just one link it wouldn't be so bad, but as you all know, you wind up with a rats nest of email chains and links. Dropbox assets, viewing links, etc. Some with passwords, some with expirations. It's not uncommon to have links flying around from 5 or 6 different services because they are all one way streets and while I may send you a Dropbox link, you send back something via ftp and hightail. With Frame.io, you create an account once (takes less than 10 seconds) and now you grant someone access to a project. When they want to access that project again, they just go to one place and remember one login. From that one place they can download full resolution assets, watch work in progress in the browser, and communicate. You can invite them to as many projects as you want and they still go to one place, and remember one password (their own). If they are Google users, they can get into Frame.io in under 3 seconds with 3 clicks, no typing, and no password to remember. Once they invest in making that first account, it makes your and their lives dramatically easier. You kill needing to send email links and every time there is an update, they can still just come back to the same pace. Nothing new to remember. On the next job, works exactly the same way. So while the initial up front investment may feel like an emotional block or nuisance, the reality it's actually not. If we want our workflows to improve we have to invest in adopting new ideas. This is not to say that in the future we may offer ways in which you can interact without logging in. We understand there are some cases where a login absolutely does not make sense and for that we have share links so you can send Vimeo style links with a password but that is view only and does not support comments.

We're also rolling out an update this week that will allow you to send a join link via email. So if you wanted to get a whole group of clients into Frame.io via email, you can send the direct join link where the project will load but a pop-up will ask them for an email. This is a bit of a more passive way of getting people to share their identity so we can keep track of what's going on and continue to make Frame.io the best collaboration platform out there! You'll see this strategy evolve over the next few months. We're doing lots of things to make getting into Frame.io easier.

 

[Terry VerHaar]

Emery, thanks for being here and letting us know what's going to happen going forward. I am curious, do you have plans to add additional features to the evaluation process, like adding metadata tags or indicating circle takes? Right now, for my needs and wants, Lightiron's LivePlay3 is the model for being able to have key collaborators add meaningful metadata to the process for editing. Would love to get a sense for the roadmap.

Thanks.

 

[Björn Benckert]

If it isn't clear from our UX and UI, we obsessively focus on simplicity. Requiring a login was a philosophical choice. Frame.io is trying to fundamentally change the workflow of emailing links. Emailing links is what makes everything suck. It's unorganized and totally impossible to keep track of. Every time you want to access something you have to find the link (and perhaps the associated password) in your email. If this were just one link it wouldn't be so bad, but as you all know, you wind up with a rats nest of email chains and links. Dropbox assets, viewing links, etc. Some with passwords, some with expirations. It's not uncommon to have links flying around from 5 or 6 different services because they are all one way streets and while I may send you a Dropbox link, you send back something via ftp and hightail. With Frame.io, you create an account once (takes less than 10 seconds) and now you grant someone access to a project. When they want to access that project again, they just go to one place and remember one login. From that one place they can download full resolution assets, watch work in progress in the browser, and communicate. You can invite them to as many projects as you want and they still go to one place, and remember one password (their own). If they are Google users, they can get into Frame.io in under 3 seconds with 3 clicks, no typing, and no password to remember. Once they invest in making that first account, it makes your and their lives dramatically easier. You kill needing to send email links and every time there is an update, they can still just come back to the same pace. Nothing new to remember. On the next job, works exactly the same way. So while the initial up front investment may feel like an emotional block or nuisance, the reality it's actually not. If we want our workflows to improve we have to invest in adopting new ideas. This is not to say that in the future we may offer ways in which you can interact without logging in. We understand there are some cases where a login absolutely does not make sense and for that we have share links so you can send Vimeo style links with a password but that is view only and does not support comments.

We're also rolling out an update this week that will allow you to send a join link via email. So if you wanted to get a whole group of clients into Frame.io via email, you can send the direct join link where the project will load but a pop-up will ask them for an email. This is a bit of a more passive way of getting people to share their identity so we can keep track of what's going on and continue to make Frame.io the best collaboration platform out there! You'll see this strategy evolve over the next few months. We're doing lots of things to make getting into Frame.io easier.

Can only say that most people working in the corporate and agency world maybe have google mail, but from testing with frame.io invites I can only tell you that most of them does not know their own password. Their sysadmin sets it all up for them on their computer and they dont use their password often enough to remember it.

Last project I sent 7 invites. 3 logged in the other four did not manage to get trough that barrier... I know silly but thats how it is. So then vimeo was used, the guys who got in really liked it so we will see if the other guys put some effort into it.

But why not make it so that I can manage to send an invite to an email and choose a password for it that then is included in the invitation email... maybe not as safe as people using their own google mail password, but then again, it´s not really top secret stuff most of us deal with even if some clients think their commericals are.

 

[Björn Benckert]

Hey everyone, Frame.io co-founder here. First off, sucks to hear many of you are having playback trouble but we are aware of the issue and we're actively working on solutions that will be available in a matter of weeks. As Chris McKechnie pointed out, we're still a brand new product/company and we've seen tremendous growth over the past few months. Building an infrastructure that supports users in over 160 countries has it challenges. Our primary data center is located on the east coast of the United States. The further you are from that data center, the greater your latency and poorer your performance will be. For many international members, this means a pretty crappy playback experience. We do use what's called a content distribution network which means when you request to play a video from a geographically distant location, we instantly move the asset from our united states server to an edge location nearest you. This type of architecture works great for a company like YouTube and a million people in Sweden all the sudden want to watch the latest viral video. It can be cached at an edge location and all future requests will be accessed from that cache. But for a service like Frame.io where videos are being accessed private and infrequently, this solution doesn't work well because it's still slow to move the file from our origin to the cahched location. So what are we going to do:

1. First, we'll offer a lower bit rate resolution (like YouTube). Right now our playback is a 5,000kbps 1080p stream. This is just too high for many connections. Offering lower bitrates will solve the problem for most of you.
2. We're designing a new architecture that can better handle the geographically dispersed users. This means if you live in Sweden you'll likely be uploading to a data center near you where that asset will permanently live. We're also testing other ideas.
3. Upload performance from regions outside the U.S. will get dramatically better in the coming weeks as well.

It's important to note that everything you upload to Frame.io get's converted server side (like YouTube and Vimeo) so it doesn't matter what you upload. That being said, resolution does matter. If you upload 1280 x 720 or lower, we will not up convert is to 1080 so the stream winds up around 2500 kbps. So that's a temporary solution to improving playback performance.

The other issue that was mentioned about it taking a while to start when you press play is a Safari bug we're aware of that we're fixing very soon. These are all just growing pains for a company that's only 4 months old. There are tons of areas we're working in improving but being able to upload and playback are obviously the most central so we're focusing on it now and this will all be a distant memory soon.

Thanks for the continued support guys!

Cool, so my guessing about mirroring was kind of right.

I know very little about this things but as it´s part of my daily work to upload things for clients that are spread out all over the place. May I ask, why don´t you guys piggy back on one of the majors like youtube or vimeo instead of hosting the files yourselves. I really like your commenting / draw with the pen feature and all but in the end of the the bandwidth, playback and download speed is allways the first priority and to me you can not have too much of them and to win that battle over youtube and vimeo with a pro dedicated system like frame.io seams like a really tough thing to conquest, no?

 

[Emery Wells]

Cool, so my guessing about mirroring was kind of right.

I know very little about this things but as it´s part of my daily work to upload things for clients that are spread out all over the place. May I ask, why don´t you guys piggy back on one of the majors like youtube or vimeo instead of hosting the files yourselves. I really like your commenting / draw with the pen feature and all but in the end of the the bandwidth, playback and download speed is allways the first priority and to me you can not have too much of them and to win that battle over youtube and vimeo with a pro dedicated system like frame.io seams like a really tough thing to conquest, no?

We plan on supporting embedding YouTube/Vimeo but there are a number of reasons we host the content ourself:

1. Complete control over security
2. Complete control over format support
3. Complete control over transcode speed. If you compare the total time to upload to YouTube and have that asset ready to view VS in Frame.io, we really destroy them. Having to always upload to YouTube/Vimeo first would not be ideal but we want to support that workflow for those who choose to work that way.
4. Ability to store original assets for download
5. Ability to support more than just video

You may not realize that Frame.io is a file sharing platform in addition to work in progress review. The original file you upload is always stored for download. You can upload formats like ProRes and preview right in the browser. You can also upload still and audio assets as well.

 

[Emery Wells]

Can only say that most people working in the corporate and agency world maybe have google mail, but from testing with frame.io invites I can only tell you that most of them does not know their own password. Their sysadmin sets it all up for them on their computer and they dont use their password often enough to remember it.

Last project I sent 7 invites. 3 logged in the other four did not manage to get trough that barrier... I know silly but thats how it is. So then vimeo was used, the guys who got in really liked it so we will see if the other guys put some effort into it.

But why not make it so that I can manage to send an invite to an email and choose a password for it that then is included in the invitation email... maybe not as safe as people using their own google mail password, but then again, it´s not really top secret stuff most of us deal with even if some clients think their commericals are.

We don't require Google but I was just pointing out how seamless that experience can be. Non Google login is almost as simple. Believe me, I hear where you are coming from and there are still a number of things we can do to make things better. Imagine if Facebook, Slack, or even Reduser did not require someone to have an identity. They wouldn't work. Identity is an important part of building a collaborative environment that fosters communication. Our vision might not be fully apparent right now but you'll continue to see things improve. It's out job to make sure everyone get's through that barrier.

 

[Emery Wells]

Emery, thanks for being here and letting us know what's going to happen going forward. I am curious, do you have plans to add additional features to the evaluation process, like adding metadata tags or indicating circle takes? Right now, for my needs and wants, Lightiron's LivePlay3 is the model for being able to have key collaborators add meaningful metadata to the process for editing. Would love to get a sense for the roadmap.

Thanks.

Lot's on the roadmap. Nothing specific to share but being able to input metadata will definitely be available in the future. We have a very flexible metadata model

 

[Emery Wells]

Cool, so my guessing about mirroring was kind of right.

I know very little about this things but as it´s part of my daily work to upload things for clients that are spread out all over the place. May I ask, why don´t you guys piggy back on one of the majors like youtube or vimeo instead of hosting the files yourselves. I really like your commenting / draw with the pen feature and all but in the end of the the bandwidth, playback and download speed is allways the first priority and to me you can not have too much of them and to win that battle over youtube and vimeo with a pro dedicated system like frame.io seams like a really tough thing to conquest, no?

Let me ask you a question. Is it really any easier for you to set someone's password on a project by project basis VS letting them create one password for themselves that will never change? The latter is most definitely easier/optimal but it's a mental/psychological block. People think creating an account is annoying. So from a product design perspective we have to figure out how to remove that barrier. The mechanics of creating an account are no more time consuming than typing in a provided password. We have some ideas we'll be testing out here. The first which you'll see this week is sending join links that anyone can click on to join a project. We actually load the project in the browser so it's visible in the background but then pop-up a box asking for an email before moving forward. It's more of a see through wall than a brick wall. Makes them feel like they are closer. We're hoping this will have a positive psychological effect.

 

[Gunleik Groven]

Just from an outside observation, but most of my clients work with many providers. Thus the link/password approach makes them feel like "not going to bed with me".

I get your approach, sort of, but I dunno if you can see this?
Then again, I do, and have not yet gone frame.io.